.png){kind=icon}
.png){kind=icon}
Return
Nonparametric changed segment detection in functional data
.png){kind=icon}
OA
AI
DOI：10.15388/namc.2026.31.44489.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We address the epidemic change point detection problem without parametric assumptions. We propose statistics based on Cram & eacute;r-von Mises-type statistic and reproducing kernel Hilbert space that iterate through all interval subsets, rescaling them to remain sensitive to both short and long epidemics. We prove limit theorems and provide quantiles for both statistics under the different parametrizations. The simulations show consistent power across a wide range of scenarios, and an application to electricity balancing prices consistently detects a market disturbance.
Keywords:
epidemic change
reproducing kernel Hilbert space
Cram & eacute
r-von Mises-type statistic
Journal
IF:
2
Papers:
52
Citations:
1.1K